Hi,
I've had user reports of the above problem - delays in posts appearing, several hours in some cases, in a specific thread. It may be connected to the 'new page' boundary i.e. a new page is forced due to number of threads.
Users are discussing it here:
http://forums.linn.co.uk/bb/showthread.p...807&page=1
And the problem thread is:
http://forums.linn.co.uk/bb/showthread.php?tid=5804
We're still using 1.4.9 (I know, I know - am hassling our guys to upgrade) but this is the first time a problem like this has occurred (or any problem for that matter, great software).
Can anyone help?
Many thanks,
olly
Can you provide a test account so I can try this out for myself. I can't really think of any reason for the delay other than maybe browsers incorrectly caching.
(2010-02-09, 11:28 AM)TimB. Wrote: [ -> ]Can you provide a test account so I can try this out for myself. I can't really think of any reason for the delay other than maybe browsers incorrectly caching.
Apparently the delay occurred late last night / early this morning, so you won't see any ill effects today. Feel free to register though - it's quick and painless and we won't spam you
Is it possible our own server was incorrectly caching pages for users?
Thanks.
are the threads moderated? that's what usually causes delays in posts appearing.
other than that, try a recount & rebuild, it won't show new page links if the threads post count is lower than the actual number.
server side caching would be unusual, client side caching would usually mean that some people see the posts whereas others don't, depending on what's in their cache.
easy method to exclude the possibility of server side cache is to edit an old visible post and see if the edit is visible or not. if it's cached the edit wouldn't show (use full edit, not quick edit, for that one)
(2010-02-09, 11:43 AM)frostschutz Wrote: [ -> ]are the threads moderated? that's what usually causes delays in posts appearing.
other than that, try a recount & rebuild, it won't show new page links if the threads post count is lower than the actual number.
server side caching would be unusual, client side caching would usually mean that some people see the posts whereas others don't, depending on what's in their cache.
easy method to exclude the possibility of server side cache is to edit an old visible post and see if the edit is visible or not. if it's cached the edit wouldn't show (use full edit, not quick edit, for that one)
No moderation and a full edit displays instantly so I'll wait to see wait our Linux guy downstairs says before I consider the rebuild.
Thanks for the input, guys.
Actually, problem appears to be 'live' i.e. users are experiencing delays in actual thread updating with post even though overviews can indicate a newer post exists.
Users seem to identify the page boundary as being key here...does anyone know if there is a reported bug with this on any version?
Thanks.